Compound tcp
This presentation is the property of its rightful owner.
Sponsored Links
1 / 22

Compound TCP PowerPoint PPT Presentation


  • 42 Views
  • Uploaded on
  • Presentation posted in: General

Compound TCP. Murari Sridharan and Kun Tan (Collaborators: Jingmin Song, MSRA & Qian Zhang, HKUST. Design goals. Efficiency Improve throughput by efficiently using the spare capacity in the network RTT fairness Intra-protocol fairness when competing with flows that have different RTTs

Download Presentation

Compound TCP

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript


Compound tcp

Compound TCP

Murari Sridharan and Kun Tan

(Collaborators: Jingmin Song, MSRA & Qian Zhang, HKUST


Design goals

Design goals

  • Efficiency

    • Improve throughput by efficiently using the spare capacity in the network

  • RTT fairness

    • Intra-protocol fairness when competing with flows that have different RTTs

  • TCP fairness

    • Must not impact performance of regular TCP flows sharing the same bottleneck

  • Stability


The compound tcp approach

The Compound TCP approach

  • Synergy between loss and delay based approaches

    • Using delay to sense network congestion

    • Adaptively adjust aggressiveness based on network congestion level

  • One flow, two components

    • Loss based component: cwnd (standard TCP Reno)

    • Scalable delay-based component: dwnd

    • TCP send window is controlled by win = cwnd + dwnd


Ctcp congestion control

CTCP congestion control

  • Vegas-like early congestion detector

    • Estimate the backlogged packets (diff) and compare it to a threshold, γ

  • Binomial increase when no congestion

  • Multiplicative decrease when loss

  • On detecting incipient congestion

    • Decrease dwnd and yield to competing flows


Ctcp congestion control1

CTCP congestion control

  • cwnd is updated as TCP Reno

  • dwnd control law

  • The above control law kicks in only when the flow is in congestion avoidance and cwnd >= 40 packets. No changes to slow start phase.


Response function

Response function


Ctcp window evolution

CTCP window evolution

dwnd = 0

dwnd

Wm

B

Ws = cwnd + dwnd

γ

W(1-b)

cwnd

D

E

F

G

Loss Free Period


Convergence and rtt fairness

Convergence and RTT fairness

  • Theorem 1: Two CTCP flows with same round trip delay converge to fair share.

  • Theorem 2: Let Th1 and Th2 present the throughput of two CTCP flows with round trip times R1 and R2, respectively. Then, the following inequality satisfied


Tcp fairness

TCP fairness

  • Bandwidth stolen

    • Let P be the aggregated throughput of m regular TCP flows when they compete with l regular flows. Let Q be the aggregated throughput when competing with high-speed flows. The bandwidth stolen by high-speed protocol flows from regular TCP flows is

  • Theorem 3: CTCP is fair and will not steal bandwidth from competing flows when , where B is the bottleneck buffer size.


Effect of gamma

Effect of Gamma

  • γ fixed at 30 packets. This works well on most

  • Delay component loses ability to detect early congestion

    • Average buffer allocated for each flow is < γ


Gamma tuning by emulation

Gamma tuning by emulation

  • Loss based component of CTCP emulates the behavior of regular TCP. The cwnd’s of competing flows converge and should be the same before hitting a packet loss.

  • At the end of every round, compute backlogged packets (Diff_reno) purely based on cwnd the loss based component.

  • On a packet loss, choose γ = 3/4 * Diff_reno. Update γ using an exponential moving average

  • Ensure γlow <= γ <= γhigh . Experimentally we have determined γlow = 5 , γhigh = 30


Implementation evaluation

Implementation & Evaluation

  • Windows platform implementation

    • Microsecond resolution RTT timer

    • Dynamic memory management for sample buffers

  • DummyNet-based Test-bed


Efficiency

Efficiency


Rtt fairness

RTT fairness


Tcp fairness effect of

TCP fairness – Effect of γ


Tcp fairness varying buffer sizes

TCP fairness – Varying buffer sizes


Intra protocol fairness and convergence of

Intra-protocol fairness and convergence of γ


Stability

Stability


Multiple bottlenecks utilization

Multiple bottlenecks – Utilization

N=400 TCP, K=50 TCP

M is either 8 CTCP or TCP flows

M flows 2.5G/10ms 1G/30ms 2.5G/10ms

N flows K flows


Multiple bottlenecks throughput

Multiple bottlenecks – Throughput

N=400 TCP, K=50 TCP

M is either 8 CTCP or TCP flows


Summary

Summary

  • CTCP is a promising approach that achieves good efficiency, RTT fairness and TCP fairness.

  • Implemented on Windows platform and verified the above properties in a range of environments.

    • Validated on test-beds, Microsoft IT high-speed links, Microsoft internal deployments, SLAC/Internet2/ESNet production links. (Refer to the “Testing TCP over Wide Area Networks” talk by Yee-Ting Lee)


Challenges

Challenges

  • In the absence of explicit feedback

    • Does delay correspond to congestion?

      • Delay indicates incipient congestion which may or may not result in packet loss.

    • Is queue build up measureable at the end systems?

      • The granularity of the RTT samples are based off high-resolution hardware timers. Helps measure variations in baseRTT.

    • Can queueing delay be more accurately estimated than end-to-end loss probability?

      • More fine grained rtt samples, loss events are rare.

    • Can we detect congestion in timescales of the order of RTT? Should we?


  • Login